Random coincidences, additional deadtime, and scatter from activity outside the field of view are potential sources of bias and noise in septa-less acquisitions from 3-D PET scanners. The purpose of this study is the characterize the magnitude of this contamination and to assess its impact on typical 3-D image data. Two phantom distributions were created to mimic contamination from activity outside the scanner in two typical imaging conditions: (I) contamination of brain studies by activity in the heart and (2) contamination of cardiac studies by activity in the bladder. These studies showed a substantial increase in random coincidences with outside activity, even for a source distant from the gantry (approximately 3O cm). Increased deadtime from the extra singles flux was also observed. However, in the data studied, there was no more than a 5% bias introduced into the reconstructed values by the outside activity. The influence of the axial location of the outside activity was also studied by several phantom distributions. It was found that beyond approximately 2O cm, the increase in scatter from outside activity is negligible. Increased randoms and deadtime remain significant out to approximately 40 cm from the gantry. The next phase of the study will be to complete these phantom studies to characterize the problem. In addition, an assessment of methods to reduce or eliminate the effects of outside activity will be made.